I. Introduction

The Seattle Metropolitan Area is growing quickly. Local giants such as the Boeing Company, Microsoft Corporation, and Amazon.com will likely continue their expansion in the area, and many new companies, especially in the tech industry, will enter the region. With this inflating economy comes an influx of residents and a greater demand for development. However, the Seattle Metropolitan Area cannot develop blindly without any consideration of the natural environment. Nature, specifically forests, is a big local selling point that makes the Pacific Northwest an attractive place to live. Therefore, the destruction of the environment in favor of development would ultimately tarnish the region’s reputation and deflate the economy.

This memo describes a systematic and rational procedure for allocating land for development. First a linear model is used to predict development demand. Second, environmentally suitable land is identified as development supply. Third, an allocation procedure is outlined with respect to development in 2020. Fourth, allocation is discussed in a specific scenario that anticipates the addition of new ferry stations around Puget Sound.

II. Predicting Development Demand

Demand is interpreted as the probability that a region will be developed by 2020. In order to calculate this probability, data from 2000 to 2010 is used to determine the factors that influence development. It is concluded that land cover type, population change, distance to highways, distance to ferry stations, and proximity to already developed land best predict new development between 2000 and 2010. The figures and tables below show the statistical significance for these variables. For example, the distance to ferry stations is on average much less for areas that became developed between 2000 and 2010 when compared to other areas.

Observations 16612
Dependent variable lc_change
Type Generalized linear model
Family binomial
Link logit
χ²(8) 451.40
Pseudo-R² (Cragg-Uhler) 0.28
Pseudo-R² (McFadden) 0.27
AIC 1220.79
BIC 1290.25
Est. S.E. z val. p
(Intercept) -2.76 0.21 -12.87 0.00
wetlands1 -0.53 1.03 -0.52 0.60
forest1 1.41 0.26 5.39 0.00
farm1 1.39 0.36 3.86 0.00
otherUndeveloped1 2.56 0.31 8.36 0.00
lagDevelopment -0.00 0.00 -6.69 0.00
pop_Change 0.00 0.00 7.12 0.00
distance_highways -0.00 0.00 -1.53 0.13
distance_ferryStations -0.00 0.00 -3.10 0.00
Standard errors: MLE

These significant variables were then used to predict changes in development between 2010 and 2020. Population change between 2010 and 2020 was estimated by extrapolating 2019 county population numbers and proportionately allocating this population growth among individual regions.

III. Assessing Supply

Supply is interpreted as the land that is environmentally suitable for development. For this calculation, two criteria for environmental sensitivity are used that identify unsuitable land. 1. The presence of sensitive land cover in 2010. Note, sensitive land is defined as wetlands and forests. 2. The presence of un-fragmented sensitive land cover patches greater than 100 acres.

Criterion 2 is a subset of criterion 1. However, the distinction is important because land fragmentation from development has particularly devastating consequences for local wildlife. For example, the construction of an access road across a stream could prevent salmon from following their customary migration patterns.

IV. Allocation

In order to properly allocate land for development, environmentally suitable sites with high demand for development must be found. The high-demand sites are represented as blue pixels in the development potential figures below. The darker the blue, the higher the development demand. However, areas with poor environmental suitability, regardless of development demand, appear light gray because the poor suitability markers are layered on top of demand. The symbology is ordered this way because it is preferable to preserve all existing patches of wetland and forest, regardless of demand.

The top figure below displays county specific allocation metrics as bar graphs. The bottom figure depicts development potential for the entire Seattle Metropolitan Area using the symbology described above.

While all of the counties initially appear to have almost no supply of environmentally suitable land, upon closer examination, opportunities for development do exist. The figure below shows Pierce County’s development potential.

Specific allocation must be considered county by county. However, we regard it as important that the constraints regarding land suitability are held constant throughout the entire region. For example, Skagit county should not independently decide that it now considers fragmentation-susceptible forests to be suitable for development. The individual counties will decide what minimum amount of demand warrants development. It is important to avoid wasting labor, time, and money by developing on land with that will be underutilized due to low development demand. However, sometimes an area with low predicted development demand in this 2020 scenario is predicted to have significantly higher demand when taking into account planned construction of new infrastructure (as discussed below).

V. Anticipating the Construction of New Ferry Stations on Puget Sound

By the end of 2020, Puget Sound will substantially expand its ferry network to increase accessibility throughout its islands and peninsulas. In the north, expanded ferry connections between Camano Island and Whidbey Island will amplify the civic capacity of Oak Harbor. The South Basin, Tacoma, and regions of the Kitsap Peninsula will also become more connected. This includes Boston Harbor, Union, and Hoodsport.

Pictured below is a 2020 predicted development demand map that anticipates these new ferry stations. Also pictured below for comparison is the original 2020 predicted development demand map.

At this scale, the differences in development demand are very fine details. However, this expansion in ferry stations shows notable new opportunities for development in Mason, Thurston, Lewis, and Skagit Counties. The figures below show development potential with and without new ferry stations factored in, as well as the projected 2010 – 2020 population change.

Both Burlington in Skagit County and Shelton in Mason County have seen significant population growth over the last 10 years. Much of the land cover in both regions is farmland, but the population growth suggests that this farmland is fit for denser redevelopment. Growth in Burlington would strengthen the area’s connection to Vancouver, British Columbia, because Burlington sits halfway between Vancouver and Seattle. Development along Interstate 5 between Lewis County and Thurston County would increase accessibility to the historic town Chehalis, the state capitol Olympia, and even Portland, Oregon.

Conclusion

The sheer amount of forest in the Seattle area limits development. Nevertheless, an intelligent and consistent land allocation plan can accommodate the anticipated influx of residents without permanently damaging the natural environment. A new network of ferries has the potential to attract developers to presently underutilized regions. However, environmental considerations must not be lost.